Model Predictive Control for Real-Time Price-Maker Bidding by Grid-Scale Battery Energy Storage Systems

Battery energy storage systems (BESS) have been increasingly adopted in power grids to support the growing renewable penetration. To participate in electricity markets, BESS need effective bidding strategies to generate revenue while inherently contributing to grid stability. However, most studies assume price-taker roles for the BESS in the electricity market despite evidence that the operation of grid-scale BESS can impact market prices. As the size of grid-scale batteries grows, conventional price-taker assumptions become less reliable. To address this, we propose a novel approach that combines a linear regression (LR) model with model predictive control (MPC) to enable real-time price-maker bidding for the BESS. The LR model captures the impact of BESS actions on electricity prices and is incorporated in a mixed-integer quadratic programming (MIQP) model, allowing the MPC algorithm to optimize bidding decisions based on predicted market dynamics. Simulations based on real data from five Chinese provinces demonstrate the model's ability to account for market influence and generate 52.7% higher arbitrage revenue compared to price-taker benchmarks. Our LR-MPC bidding strategy facilitates effective participation of the BESS from price-takers to price-makers by addressing the growing influence of large-scale BESS in the electricity market.